METHODS: Relevant randomised controlled trials were searched using appropriate keywords in health-related databases. We performed network meta-analysis within a frequentist framework. The endpoint assessed was the R0 resection rate. We assessed consistency and transitivity assumptions that are necessary for network meta-analysis. This study only used data from published studies. The need for consent from participants was waived by the Ethics Review Committee of the International Medical University in Malaysia.

RESULTS: Six randomised controlled trials involving 1700 patients were identified. A network plot was formed with five neoadjuvant regimens [DLX (pyrimidine analogue + platinum compounds + chemoradiotherapy), DELX (pyrimidine analogue + epipodophylllotoxins/etoposide + platinum compounds + chemoradiotherapy), ADL (anthracycline + pyrimidine analogue + platinum compounds), ADM (anthracycline+ pyrimidine analogue + anti-folate compounds) and LTX (platinum compounds + taxane + chemoradiotherapy)] and surgery alone for management of patients with advanced gastroesophageal and gastric cancer. Assumptions required for a network meta-analysis such as consistency ((global test: Chi2 (1): 3.71; p:0.054)), and the transitivity in accord to the characteristics of interventions considered in this review were not violated. In the network comparison, surgery alone has a lower R0 resection rate compared with LTX (OR 0.2, 95%CI:0.01, 0.38) or DLX (OR 0.48, 95%CI: 0.29, 0.79). LTX has higher resection rate compared with DLX (OR 2.47, 95%CI: 1.08 to 5.63), DELX (OR 106.0, 95%CI: 25.29 to 444.21), ADM (OR 5.41, 95%CI: 1.56 to 18.78) or ADL (OR 3.12, 95%CI: 1.27 to 7.67). There were wide or very wide CIs in many of these comparisons. Overall certainty of the evidence was low or very low. Further research in this field is very likely to have an important impact on our confidence in the R0 resection rates between LTX versus other neoadjuvant chemotherapy is likely to change the estimate.

METHODS AND MATERIALS: Patients with GCTs and treated with at least two cycles of BEP chemotherapy between January 2003 and Oct 2009 were eligible for this study. Patients received 4-6 cycles of bleomycin 30,000IU IV D1, D8 and D15 and either etoposide 100mg/m2 IV D1- D5 and cisplatin 20mg/m2 IV D1- D5 (5 day BEP regimen) or etoposide 165 mg/m2 D1- D3 and cisplatin 50mg/m2 D1-3 (3 day BEP regimen) every three weeks per cycle. All patients received prophylactic granulocyte colony-stimulating factor (GCSF) from days 6 to 10 of each cycle. The overall response rates, 2 year progression-free survival and overall survival of the whole cohort were assessed.

RESULTS: Thirty patients fulfilled the inclusion criteria. Non-seminomatous GCTs comprised 93.3% of cases and gonadal and mediastinal primary sites were the most common. Sixty percent were classified as IGCCCG poor risk disease. Median follow-up was 26.6 months. The overall response rate (CR+PR) was 70%. The two year PFS and OS were 70% and 66%. There was a significant difference in terms of the overall response rate (85% vs 40%, p = 0.03) and in PFS (94.7% vs 50%, p = 0.003) between gonadal and extragonadal primary sites.